{"paper":{"title":"Seconds-Aligned PCA-DAC Latent Diffusion for Symbolic-to-Audio Drum Rendering","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"A latent diffusion model predicts principal-component coordinates of a frozen audio codec to render drum audio from symbolic timing with better spectral and transient accuracy than regression.","cross_cats":[],"primary_cat":"cs.SD","authors_text":"Dimos Makris, Konstantinos Soiledis, Konstantinos Tsamis, Maximos Kaliakatsos Papakostas","submitted_at":"2026-05-13T11:59:41Z","abstract_excerpt":"Symbolic-control drum generation requires preserving explicit event timing and dynamics while synthesizing acoustically plausible waveforms. We present Sec2Drum-DAC, a conditional latent-diffusion model for symbolic-to-audio drum rendering. The model conditions on event features sampled in physical time at codec-frame locations and predicts standardized principal-component coordinates of frozen DAC summed-codebook embeddings rather than waveform samples. In the evaluated DAC configuration, 72 principal components capture the observed training-frame summed-latent subspace under the stated SVD t"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"Across 1,733 held-out four-beat windows, PCA diffusion improves paired spectral and transient metrics over deterministic PCA regression and a symbolic rendering baseline, while direct regression remains stronger on phase-sensitive waveform L1.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"That the 72 principal components derived from training data via SVD threshold sufficiently represent the variations needed for high-quality reconstruction of held-out drum audio when decoded through the frozen DAC.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"Sec2Drum-DAC renders drum audio from symbolic inputs via diffusion on PCA-reduced DAC latents, improving spectral and transient metrics over regression baselines on 1733 held-out windows.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"A latent diffusion model predicts principal-component coordinates of a frozen audio codec to render drum audio from symbolic timing with better spectral and transient accuracy than regression.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"e7baf641c8e43eb57167bff6ec85ff68b330d1434a56b9d15cdbe6ec86c83b2f"},"source":{"id":"2605.13404","kind":"arxiv","version":1},"verdict":{"id":"5091d3f7-5eae-4fe9-8d0c-1c5d12760882","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-14T18:34:08.695480Z","strongest_claim":"Across 1,733 held-out four-beat windows, PCA diffusion improves paired spectral and transient metrics over deterministic PCA regression and a symbolic rendering baseline, while direct regression remains stronger on phase-sensitive waveform L1.","one_line_summary":"Sec2Drum-DAC renders drum audio from symbolic inputs via diffusion on PCA-reduced DAC latents, improving spectral and transient metrics over regression baselines on 1733 held-out windows.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"That the 72 principal components derived from training data via SVD threshold sufficiently represent the variations needed for high-quality reconstruction of held-out drum audio when decoded through the frozen DAC.","pith_extraction_headline":"A latent diffusion model predicts principal-component coordinates of a frozen audio codec to render drum audio from symbolic timing with better spectral and transient accuracy than regression."},"references":{"count":33,"sample":[{"doi":"","year":1906,"title":"Takuya Akiba, Makoto Shing, Yujin Tang, Qi Sun, and David Ha","work_id":"4acd1abb-5322-43c7-b701-6718c3d52b72","ref_index":1,"cited_arxiv_id":"1906.02569","is_internal_anchor":true},{"doi":"","year":2023,"title":"MusicLM: Generating Music From Text","work_id":"15e6566e-1c36-468f-966e-823248cbf87f","ref_index":2,"cited_arxiv_id":"2301.11325","is_internal_anchor":true},{"doi":"","year":2016,"title":"madmom: A new Python audio and music signal processing library","work_id":"d95aeaf7-c243-4347-80f8-87171845aa3a","ref_index":3,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2023,"title":"AudioLM: A language modeling approach to audio generation.IEEE/ACM Transactions on Audio, Speech, and Language Processing, 31:2523–2533, 2023","work_id":"9f88e30d-7bc1-4391-b8f2-d0925e784600","ref_index":4,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2026,"title":"DARC: Drum accompaniment generation with fine-grained rhythm control","work_id":"8124ea20-a88f-4d9b-bee4-834478e9d0bf","ref_index":5,"cited_arxiv_id":"","is_internal_anchor":false}],"resolved_work":33,"snapshot_sha256":"2a2bc55b4e087647d2f6c756741672d55372f60b3efb4bdc05bc3543f169e799","internal_anchors":4},"formal_canon":{"evidence_count":1,"snapshot_sha256":"a516c148e5b2ab5fc5d787375119a8634b20f4cd815ff8ff929da5bf4ed12a66"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"}